WiseDiag Skin

When It Helps

Use this skill when you already have a local skin photo and want the model to go beyond generic image description: it can point to possible skin conditions, give a brief analysis, and produce recommendations. It is useful for engineering workflows where the input is a local file path, and the script handles upload, API calls, streaming output, and saving the final report without hand-rolled multipart/form-data handling.

How It Works

The workflow is local binary upload + WiseDiag AI analysis + streaming output:
- Pass a local image with -f or --file, such as skin.jpg or rash.png;
- The script reads WISEDIAG_API_KEY from the environment and calls the WiseDiag service;
- Optionally refine the request with --question; by default it asks for possible conditions;
- Execution can take 30–120 seconds, especially for large images, and the script prints streaming output;
- The final result is saved automatically to ~/.openclaw/workspace/WiseDiag-Skin/{name}.md.

Limits And Notes

The output is for reference only and is not a medical diagnosis. Images are uploaded to WiseDiag cloud servers for processing, so avoid uploading photos that contain sensitive or private content. Use supported formats such as JPG, PNG, WebP, GIF, BMP, or TIFF, and keep the photo clear and well-lit. If you see WISEDIAG_API_KEY is not set, authentication failure, or File not found, check the environment variable, API key status, and local path respectively.